Skip to main content
The Globe and Mail
Support Quality Journalism
The Globe and Mail
First Access to Latest
Investment News
Collection of curated
e-books and guides
Inform your decisions via
Globe Investor Tools
per week
for first 24 weeks

Enjoy unlimited digital access
Enjoy Unlimited Digital Access
Get full access to
Just $1.99 per week for the first 24 weeks
Just $1.99 per week for the first 24 weeks
var select={root:".js-sub-pencil",control:".js-sub-pencil-control",open:"o-sub-pencil--open",closed:"o-sub-pencil--closed"},dom={},allowExpand=!0;function pencilInit(o){var e=arguments.length>1&&void 0!==arguments[1]&&arguments[1];select.root=o,dom.root=document.querySelector(select.root),dom.root&&(dom.control=document.querySelector(select.control),dom.control.addEventListener("click",onToggleClicked),setPanelState(e),window.addEventListener("scroll",onWindowScroll),dom.root.removeAttribute("hidden"))}function isPanelOpen(){return dom.root.classList.contains(}function setPanelState(o){dom.root.classList[o?"add":"remove"](,dom.root.classList[o?"remove":"add"](select.closed),dom.control.setAttribute("aria-expanded",o)}function onToggleClicked(){var l=!isPanelOpen();setPanelState(l)}function onWindowScroll(){window.requestAnimationFrame(function() {var l=isPanelOpen(),n=0===(document.body.scrollTop||document.documentElement.scrollTop);n||l||!allowExpand?n&&l&&(allowExpand=!0,setPanelState(!1)):(allowExpand=!1,setPanelState(!0))});}pencilInit(".js-sub-pencil",!1); // via darwin-bg var slideIndex = 0; carousel(); function carousel() { var i; var x = document.getElementsByClassName("subs_valueprop"); for (i = 0; i < x.length; i++) { x[i].style.display = "none"; } slideIndex++; if (slideIndex> x.length) { slideIndex = 1; } x[slideIndex - 1].style.display = "block"; setTimeout(carousel, 2500); }

Build, Borrow or Buy: Solving the Growth Dilemma

Laurence Capron and Will Mitchell/Harvard Business Review Press

The following is an edited excerpt from Build, Borrow, or Buy: Solving the Growth Dilemma by Laurence Capron and Will Mitchell. Reprinted by permission of Harvard Business Review Press. Copyright 2012 Harvard Business School Publishing Corporation. All rights reserved.

Internal Exploratory Environments Important Routes to Novel Internal Changes

When your firm lacks the relevant skills or organization to develop a targeted resource, an internal exploratory environment can generate information that may help resolve the uncertainties of a promising project and even lead to a breakthrough. In such an environment, people can experiment with projects radically different from their ongoing activities, or an independent unit is created to develop new resources outside the mainstream organization. Internal exploratory environments encompass both small-scale Skunk Works projects and larger-scale semi-autonomous units. These exploratory environments are often appropriate when there is no viable external source for a strategically important resource.

Story continues below advertisement

The general manager of an online business within a European company told us, "When we cannot acquire new resources on labour markets or from other firms, we experiment internally. We experiment, fail, and learn. We develop the strategy ourselves, by refining our practices within a very loose framework ."

Let's look at the two main variants of this approach.

Skunk Works projects are small-scale undertakings within an active operating unit. Such projects typically consist of one or a few people who work together on experiments to create new goods, services, processes, business models, or other innovations. Skunk Works projects are done in addition to the members' full-time jobs. Some projects are informal nights-and-weekends efforts that fly under the corporate radar and often divert resources from other projects, while others are active parts of a company's formal technology strategy.

Skunk Works projects have met with various levels of failure and success. During the early 1960s, Frito-Lay executive Arch West reportedly diverted money from other budgets for quiet research on a new product line, which eventually became Doritos. Recognizing the creative potential of Skunk Works, businesses in growing numbers actively encourage employees to spend a portion of their time on self-driven experimentation. Skunk Works allow for small-scale explorations in which failures can be shut down easily and projects that show potential earn their way to increasing investments.

There are many examples of successful Skunk Works as formal parts of technology strategy:

• 3M has long encouraged its people to explore innovations as part of their ongoing work; this practice helps the company continually refresh its product line.

• Google asks that staff members spend 20 per cent of their time investigating independent ideas; the company estimates that as many as half its new products (including Gmail, Google News, Orkut, and AdSense services) come from this Skunk Works time.

Story continues below advertisement

• Projects from Shell's "Game Changer" innovation model have ranged from new extraction technologies to new ocean drilling techniques to hydrogen fuel cell distribution models that would radically change the petroleum-based business model.

Semi-autonomous units are larger-scale ventures in which a company creates a separate operating unit that is typically empowered to draw resources from established units. Unlike Skunk Works, which may be formal or informal, semi-autonomous units almost always require a formal charter. But they resemble Skunk Works in that they allow for experimentation away from the core. This degree of freedom limits the risk that a firm's existing activities and pressures could constrain innovation efforts that rely on highly unconventional methods. When semi-autonomous units fail, as they often do, they are easy to shut down or sell. When they succeed, their successes can be integrated into the company. Some successes are so significant; they change the direction of a company.

When Hewlett-Packard decided, in the 1970s, to experiment with PC laser and inkjet printers, it considered buying several companies with a printer technology base. After assessing the targets carefully, HP decided they were overpriced relative to the strength of their technologies.

But the company did not want to build a new printer business internally. The business model for printers would have conflicted strongly with the incumbent businesses: new design timelines and criteria would clash with customary ways of evaluating and providing resources for new projects, and printer sales channels would differ substantially from those serving HP's minicomputers.

Quite simply, the PC printer business could have been squashed by the established divisions before getting off the ground. So, HP set up a semi-autonomous unit, based in Idaho and Washington State, well away from the corporate core in California.

HP didn't leave the new business entirely on its own. It assigned leadership of the initiative to a highly influential senior executive who reported directly to the board and was empowered to obtain whatever resources the unit needed to grow. The company possessed some of the relevant skills in electronics and assembly for designing new printers, but it lacked other key resources (most notably, toner technology). So the fledgling printer unit developed a focused partnership with Canon, the global leader in toner technology. Once the printer business was a demonstrated success, HP gradually recentered its business around printers, eventually selling off its once-core scientific instruments business.

Story continues below advertisement

Similarly, IBM set up a semi-autonomous business unit – in Florida, far from its New York headquarters – to launch its new PC business. The company's resource strength was high, with strong relevant skills in electronics and assembly. These skills would allow it to create a new type of computer more effectively than competitors could. But the organizational fit was low. IBM sold "heavy iron" mainframe computers predominantly to business enterprises. The PC would begin as a consumer product requiring a substantially different business model. (Businesses didn't purchase PCs in volume until the mid- to late 1980s.) Compared with IBM's expensive mainframes, the PC was a small-ticket item that would call for new sales channels. But once the PC business demonstrated enough success to quell internal resistance, IBM absorbed it back into the core company.

Like any other high-risk or speculative initiative, semi-autonomous units can fail, sometimes exacting a high cost. General Motors created the Saturn division during the 1980s as a means of experimenting with new ways to assemble and sell cars. The division initially succeeded in both design quality and market acceptance. But strong internal and external pressure from traditional stakeholders (company leadership, representatives, dealerships, and suppliers) hampered GM's efforts to integrate Saturn's successful innovations back into the core company. The same organizational pressures that had led GM to launch Saturn as a semi-autonomous unit ultimately impeded GM's ability to integrate the successes from the experiment.

Report an error Editorial code of conduct
Tickers mentioned in this story
Due to technical reasons, we have temporarily removed commenting from our articles. We hope to have this fixed soon. Thank you for your patience. If you are looking to give feedback on our new site, please send it along to If you want to write a letter to the editor, please forward to

Welcome to The Globe and Mail’s comment community. This is a space where subscribers can engage with each other and Globe staff. Non-subscribers can read and sort comments but will not be able to engage with them in any way. Click here to subscribe.

If you would like to write a letter to the editor, please forward it to Readers can also interact with The Globe on Facebook and Twitter .

Welcome to The Globe and Mail’s comment community. This is a space where subscribers can engage with each other and Globe staff. Non-subscribers can read and sort comments but will not be able to engage with them in any way. Click here to subscribe.

If you would like to write a letter to the editor, please forward it to Readers can also interact with The Globe on Facebook and Twitter .

Welcome to The Globe and Mail’s comment community. This is a space where subscribers can engage with each other and Globe staff.

We aim to create a safe and valuable space for discussion and debate. That means:

  • Treat others as you wish to be treated
  • Criticize ideas, not people
  • Stay on topic
  • Avoid the use of toxic and offensive language
  • Flag bad behaviour

Comments that violate our community guidelines will be removed.

Read our community guidelines here

Discussion loading ...

To view this site properly, enable cookies in your browser. Read our privacy policy to learn more.
How to enable cookies